Open-Source Quadcopter for Autonomous UAV Networks for Surveillance using Zigbee Radios and Thermal Imaging

This paper details the development of an open-source quadcopter prototype for UAV-to-UAV communication via a Zigbee network. The prototype integrates a thermal imaging system for surveillance and autonomy to fly to destinations determined onboard without ground control. The prototype is assembled on a Parrot Anafi AI quadcopter utilizing a Raspberry Pi Zero 2 W, an XBee transciever, and a TOPDON TC001 thermal camera. UAV operations are executed onboard via services of the Air SDK open-source architecture and implemented in Python and C++. Ground control of the UAV runs on a separate WiFi network, and it overrides the autonomous operation for safety. Flight tests demonstrate successful UAV-to-UAV communication links of up to 113 meters long in an urban environment. Periodic acquisition of thermal images along with corresponding telemetry data (i.e., altitude, geo-location, pitch, yaw, roll) were confirmed. A separate service controls the flight trajectory, which can be determined as a result of onboard image and telemetry data processing in coordination with other equally equipped UAVs.